Skip to content

README.md

Latest commit

 

History

History
100 lines (68 loc) · 4.04 KB

README.md

File metadata and controls

100 lines (68 loc) · 4.04 KB

Case Study Datasets

Three expert opinion datasets from Czech public policy, collected during COVID-19 pandemic planning. Each demonstrates a different aspect of the BeCoMe method.

Dataset Experts Type Key Feature
budget_case.txt 22 (even) Fuzzy intervals Moderate consensus
floods_case.txt 13 (odd) Fuzzy intervals Polarized opinions
pendlers_case.txt 22 (even) Likert scale Crisp values

File Format

CASE: CaseName
DESCRIPTION: Scenario description
EXPERTS: N

# Format: ExpertID | Lower | Peak | Upper
Expert1 | 10 | 15 | 20
Expert2 | 12 | 18 | 25

Lower, Peak, Upper form a fuzzy triangular number (pessimistic, most likely, optimistic). Constraint: Lower ≤ Peak ≤ Upper. For Likert scale data, all three values are equal.

Datasets

Budget Case

Government officials estimated COVID-19 budget support needs (0-100 billion CZK). The 22-member panel included deputy ministers from multiple ministries, the Police President, Fire Rescue Director, and Chief Hygienist.

Opinions varied by ministry priorities — Interior proposed higher support (60-90 billion), Education suggested lower (15-40 billion). With an even expert count, the median calculation averages two middle values.

Chairman | 40 | 70 | 90
Deputy Minister of MI | 60 | 90 | 90
Deputy Minister of MEYS | 15 | 40 | 60

Floods Case

Experts debated what percentage of arable land should be converted for flood prevention. The 13 participants — hydrologists, land owners, rescue coordinators, economists — held sharply divided views.

Hydrologists recommended 37-47% reduction. Land owners proposed 0-4%. This polarization makes the Floods case useful for demonstrating BeCoMe's outlier resistance. With an odd expert count, the median is simply the middle element after sorting.

Hydrologist 1 | 37 | 42 | 47
Land owner 1 | 2 | 3 | 4
Land owner 2 | 0 | 0 | 2

Pendlers Case

Officials rated whether cross-border commuters should be allowed to travel during pandemic restrictions. Unlike other cases, responses used a Likert scale: 0 (strongly disagree), 25, 50, 75, 100 (strongly agree).

Foreign Affairs strongly supported travel (100). Interior and Defense opposed it (0). The majority leaned toward disagreement — 11 of 22 chose "rather disagree" (25). Likert values are crisp: Lower = Peak = Upper.

Deputy Minister of MFA | 100 | 100 | 100
Deputy Minister of MI | 0 | 0 | 0
Deputy Minister of MF | 50 | 50 | 50

Loading Data

from examples.utils.data_loading import load_data_from_txt

opinions, metadata = load_data_from_txt("examples/data/budget_case.txt")

print(metadata["case"])        # "Budget"
print(len(opinions))           # 22
print(opinions[0].expert_id)   # "Chairman"
print(opinions[0].opinion)     # FuzzyTriangleNumber(40, 70, 90)

Data Quality

All datasets validated for:

  • Format compliance (header structure, pipe-separated values, UTF-8)
  • Fuzzy number constraint (Lower ≤ Peak ≤ Upper)
  • Numerical precision against original Excel calculations (tolerance: 0.001)

Expected results stored in ../../tests/reference/. Integration tests in ../../tests/integration/test_excel_reference.py.

Provenance

Expert opinions collected during COVID-19 pandemic emergency planning in the Czech Republic. Identifiers preserve roles (e.g., "Deputy Minister of MI") while anonymizing individuals.

Limitations

  • Small samples (13-22 experts) — not large-scale surveys
  • Czech policy context during COVID-19 — may not generalize
  • Government officials and emergency services — specific perspective
  • Floods case reflects stakeholder conflict, not objective risk assessment

Related Documentation